| 00:46:03 | kergoth | whats the cmdline opt for busybox sh (ash specifically) to spawn a login shell? |
| 00:47:37 | BZFlag | busybox -sh should work. |
| 00:49:08 | kergoth | hmm k |
| 00:49:25 | kergoth | does busybox sh support loading /etc/profile and ~/.profile? |
| 00:49:27 | BZFlag | kergoth: looks like that's if for ash anyway. no option, just the -sh format. |
| 00:49:40 | kergoth | ah |
| 00:50:03 | BZFlag | yes. /etc/profile first, then .profile looks like .profile might assume you are in $HOME first. |
| 00:50:06 | kergoth | if i call /bin/sh, where sh is a link to busybox .. so that argv[0] is sh.. will it behave the same? |
| 00:50:09 | kergoth | ah, k |
| 00:50:52 | BZFlag | have not tried. you might need to exec() where you can pass it argv[0] as well as the path to the binary. |
| 00:51:18 | kergoth | hmm k |
| 00:51:24 | BZFlag | a "--login" would be nice, huh? |
| 00:51:28 | kergoth | precisely :-) |
| 00:52:17 | BZFlag | patches welcome. ;-) |
| 00:52:38 | kergoth | yeah yeah ;-) |
| 01:34:13 | | wish i knew, jacques_gone |
| 01:34:13 | jacques_gone | ibot: what? |
| 01:34:26 | | i don't know, jacques |
| 01:34:26 | jacques | ibot: huh? |
| 02:42:48 | | i guess everyone is Disconnect |
| 02:42:48 | jacques | ibot: where is everyone? |
| 02:43:01 | | jacques: excuse me? |
| 02:43:01 | jacques | ibot: you make no sense |
| 02:43:12 | | ...but ibot is <reply> a lonely little bot with too much to remember and not enough spinach to eat. or a dork or a dumbo or not very swift... |
| 02:43:12 | jacques | ibot: you are not making any sense |
| 02:43:25 | | okay, jacques. |
| 02:43:25 | jacques | ibot: you are also not making any sense |
| 04:27:25 | | privet, Hiraghm |
| 04:27:25 | Hiraghm | hi |
| 04:27:44 | Hiraghm | anybody home? |
| 06:10:37 | BZFlag | hi chad |
| 06:11:08 | BZFlagZaurus | hi timmy boy |
| 06:11:14 | BZFlag | hey watch it |
| 06:11:57 | BZFlagZaurus | oh yeah, ya think so eh... |
| 06:12:19 | DarkDream | BZFlag: YOU HAVE A ZAURUS ? :)))))))))))))))))))))))))))))))))))))))))))))))) |
| 06:12:20 | BZFlagZaurus | how bout them Jazz! |
| 06:12:24 | DarkDream | Is it really great ? :) |
| 06:13:01 | BZFlagZaurus | Its awesome! |
| 06:13:05 | BZFlag | DarkDream: yeah. |
| 06:13:12 | BZFlag | my friend is over here playing. |
| 06:13:32 | DarkDream | I plane to buy one in september or october :) |
| 06:13:45 | DarkDream | Maybe the price will be cheaper |
| 06:13:46 | DarkDream | :) |
| 06:14:00 | BZFlag | I'm the zaurus linux kernel maintainer. http://www.arm.linux.org.uk/developer/machines/ |
| 06:16:45 | BZFlagZaurus | Just get the Nav... it should be out for Christmas, but we cant say much about it ... confidential and all. |
| 06:17:20 | DarkDream | BZFlag: it's another PDA on linux ? |
| 06:17:36 | DarkDream | In fact I work for Sony and our PDA are on Palm OS |
| 06:17:39 | DarkDream | grrrrrrrrrrrrrrrrr :( |
| 06:18:21 | DarkDream | I've to leave to work |
| 06:18:26 | DarkDream | But it"s a nice place :) |
| 06:18:30 | DarkDream | I will come |
| 06:18:31 | DarkDream | ;) |
| 06:19:10 | BZFlag | Dark-Work: zaurus is linux. vercel will be linux too. |
| 06:30:23 | jacques | BZFlag: you have jtag now? |
| 06:31:13 | BZFlag | not yet |
| 06:37:19 | jacques | ah ok I misread the topic |
| 13:03:27 | scanline | is away: still studying... |
| 13:35:46 | RedBeard | 'morning |
| 13:36:20 | sjhill | g'morning |
| 13:36:28 | TomW | Yes, it is a beautifull moring in the Pocono Mountains of Pennsylvania! |
| 13:36:37 | sjhill | misses PA |
| 13:36:38 | TomW | s/moring/morning/ |
| 13:37:09 | sjhill | TomW: i grew up in Ohio about 40 minutes from the OH/PA border |
| 13:37:27 | TomW | It is 62 degress, 34% humidity, and a gentle breeze finding it's way in through the open window. Everything is so green this morning! |
| 13:38:01 | TomW | sjhill: I am in NorthEaster PA, near the Deleware Water Gap. |
| 13:38:34 | TomW | sjhill: where are you now? |
| 13:38:43 | sjhill | TomW: Cedar Rapids, Iowa |
| 13:38:58 | sjhill | parents still live near East Palestine, OH |
| 13:40:11 | sjhill | i love Iowa, but i enjoy Appalachia' too |
| 13:40:23 | TomW | sjhill: It's okay now, you can come back home. The Govenor has been running adds on PA television about how many high Tech jobs we have in this state. So, you can take advantage of the Amnesty to return home. :-D |
| 13:40:34 | sjhill | lol |
| 13:40:46 | TomW | heh |
| 13:40:53 | sjhill | i almost took a job with TimeSys |
| 13:41:58 | TomW | there is one add with your "fairy job-father" who shows this young man what he can afford in the "wast" (a crappy one room apartment) then he shows him what he can afford here (a spacious loft)... |
| 13:42:40 | sjhill | there's truth to that too |
| 13:43:17 | TomW | Allentown has turned into a software area, Scranton has a lot of light electronics manufacturing, Philadelphia has a lot of MIS. |
| 13:43:44 | TomW | Govenor Casey's work is really starting to pay off with jobs now. |
| 13:44:08 | sjhill | how's Pittsburgh these days? |
| 13:44:38 | TomW | When was the last time you saw Pittsburgh? |
| 13:44:53 | sjhill | mmm....about 2-1/2 years ago |
| 13:45:07 | sjhill | maybe 3 |
| 13:45:22 | TomW | Ok, then you have one on me, I was out there about 5 years ago and was amazed at how much that town had changed! |
| 13:46:46 | RedBeard | sjhill: I |
| 13:46:59 | TomW | I grew up in Northwestern New Jersey, a stone's throw from PA. I moved here over 15 years ago, I love this state. Level headed people here. |
| 13:46:59 | sjhill | RedBeard: 'I' what? :) |
| 13:47:07 | RedBeard | 've heard that timesys is a bit nosy in their employment interviews |
| 13:47:34 | RedBeard | as in asking a lot of questions of the individual that they wouldn't get answered by asking an former employer |
| 13:47:56 | RedBeard | such as last performance review, and last salary to the point of requiring a copy of the last w2 |
| 13:48:11 | sjhill | RedBeard: i wouldn't give them any of that information |
| 13:48:24 | sjhill | RedBeard: thx for the info, hopefully that's changed though |
| 13:48:30 | RedBeard | yeah, that's what the person that was telling me about it was saying too |
| 13:48:32 | sjhill | RedBeard: how recent did you hear that? |
| 13:48:44 | RedBeard | sjhill: I doubt it has changed, I was hearing about this yesterday |
| 13:48:56 | sjhill | sheesh |
| 13:49:47 | RedBeard | alhaz was telling me, he was asked those things in an email interview in the last few days |
| 13:49:55 | sjhill | TomW: true...i think that can be said of most MidWestern people...except maybe West Virginians :) |
| 13:51:40 | RedBeard | I've a general programing question for you guys, how do I step through a program, line by line, as in debug type things |
| 13:51:41 | RedBeard | ? |
| 13:52:10 | TomW | on an X86 box, or on the Tux? |
| 13:52:17 | RedBeard | x86 |
| 13:52:25 | TomW | use gdb. |
| 13:52:32 | RedBeard | k, thanks |
| 13:53:02 | TomW | next |
| 13:53:10 | TomW | next will step you. |
| 13:53:18 | RedBeard | I couldn't remember what it was called, I've used it a little bit |
| 13:53:22 | RedBeard | next? okay |
| 13:53:56 | TomW | I occasionally use gdb for the tough problems, usually printf works for me (or printk). |
| 13:54:20 | TomW | now, I got one for you, ever use ioremap? |
| 13:54:39 | RedBeard | TomW: no, I don't know what ioremap is, I've never heard of it |
| 13:55:22 | RedBeard | I had a bunch of printf statements in this code last night, it was a mess, but I may just have been too tired |
| 13:55:30 | RedBeard | TomW: what is ioremap? |
| 13:55:33 | TomW | I have been writing a simple device driver to wiggle CS3 on the Tux, and been having the devil's own time of getting CS3 to move. I am about to start tracing through the calls inside the kernel to see what the hell is happending. |
| 13:56:06 | RedBeard | ahh, I'm not ready to look at that kind of problem |
| 13:56:35 | TomW | according to RMK (Russell), we should be using ioremap to map the physical hardware over to a virtual address so that we can access it from inside the kernel device drivers. |
| 13:57:27 | TomW | <sigh> I guess I will ave to fire up the email client and risk sounding stupid by posting to linux-arm-kernel then.. |
| 13:59:23 | TomW | The "Linux Device Drivers" book doesn't cover accessing I/O in much detail, they glaze over the real meat. |
| 13:59:37 | RedBeard | okay, here's a stupid question, how/what do I use to send output to stderr? |
| 13:59:55 | RedBeard | printf(sdterr, "some text"); |
| 13:59:58 | RedBeard | didn't do it |
| 14:00:06 | TomW | write (stderr, "foo"); |
| 14:00:15 | RedBeard | but that's what it looked like should from the man page |
| 14:00:16 | RedBeard | okay |
| 14:00:19 | RedBeard | thanks! |
| 14:01:24 | TomW | write (STDERR, "foo"); |
| 14:01:56 | RedBeard | TomW: does write() accept printf style formating? write(STDERR, "this is an int %d", someint) |
| 14:02:10 | RedBeard | nm, I'll read the man page |
| 14:02:16 | TomW | you can use numerics for the channel: write (0, "foo") goes to stdout, write (1, "foo") goes to stderr. IIRC. |
| 14:02:24 | RedBeard | oh, cool |
| 14:02:50 | TomW | Yeah, gets you started until you figure out which header STDERR is defined in. |
| 14:04:00 | TomW | write (1, prompt, strlen (prompt)); |
| 14:04:18 | RedBeard | cool, thanks much |
| 14:04:20 | TomW | ... so I still use the numbers. |
| 14:04:35 | TomW | bad form, but.. |
| 14:05:01 | TomW | getting dynamic input from the keyboard is a bit of a pain.. |
| 14:07:52 | TomW | You want the unbuffered keyboard read stuff? |
| 14:09:21 | RedBeard | no, I'm just trying to get the output to stdout (working just fine) and the debug messages I've added in, to go to stderr |
| 14:09:30 | RedBeard | or a seperate file would be fine I guess |
| 14:10:33 | TomW | then open a file stream: FILE foo; foo = fopen ("my-file.txt", "w"); printf (foo, "this is some text"); fclose (foo); |
| 14:10:49 | TomW | s/printf/fprintf/ |
| 14:11:14 | RedBeard | ahh, that'd be clean |
| 14:11:15 | RedBeard | thanks |
| 14:12:03 | TomW | RedBeard: haven't done much C programming under MSDOS, eh? File streams on DOS were copies of the unix streams. |
| 14:12:15 | RedBeard | I haven't done much programing period |
| 14:12:31 | TomW | heh, well, all it takes is imagination. |
| 14:12:34 | RedBeard | and what I have done was several years ago and mostly in pascal (borland turbo pascal at that) |
| 14:13:14 | RedBeard | I'm having a lot of fun working through simple excersizes, but I'm finding the current excersize a bit tough |
| 14:13:33 | RedBeard | I'm sure I'm just missing something simple at this point but it's frustrating. |
| 14:13:36 | TomW | I like pascal, I did several years of UCSD Pascal on the Apple II (many years ago). Then I did / do Delphi, I like Object Pascal better than I like C++. |
| 14:13:58 | TomW | C++ sucks. |
| 14:14:47 | RedBeard | I'm trying to get my c programing skills up to a minimum usable standard, for job purposes |
| 14:14:52 | RedBeard | or, job hunt |
| 14:15:15 | RedBeard | and I've wanted to program for quite a while, I've got the time to learn, I'm having fun learning... |
| 14:16:21 | TomW | yeah, now write something useful. Or not, whatever the case. I have a lot of stuff that I wrote that had no purpose other than to learn how to do something, or sketch out an idea, or scratch an itch. |
| 14:16:49 | RedBeard | yeah, I'm working toward writing useful stuff |
| 14:18:11 | TomW | I haven't done much C programming of little "get it done" utilities since I discovered perl. I write a lot of stuff in perl now. |
| 14:18:29 | RedBeard | I can't read the perl I wrote 5 years ago |
| 14:20:28 | TomW | to answer your "write(STDERR, "this is an int %d", someint)" question, no. write only prints out strings of data. You would have to do something like "char it[256]; sprint (it, "this is an int %d", someint); write (0, it, strlen(it)): " |
| 14:20:44 | TomW | s/sprint/sprintf/ |
| 14:20:52 | RedBeard | okay |
| 14:20:55 | RedBeard | thanks |
| 14:22:05 | TomW | Which editor are you using? |
| 14:22:11 | RedBeard | vim |
| 14:22:33 | RedBeard | with syntax coloring enabled (using the evening color scheme) |
| 14:23:34 | TomW | put your cursor on a library function name, like sprintf, then type a capital K, it will bring up the manpage on that function. |
| 14:23:45 | RedBeard | oh, cool! |
| 14:23:47 | RedBeard | thanks |
| 14:24:13 | RedBeard | that's sweet |
| 14:24:30 | TomW | I stumbled on that one day, meant to hit 'k' to go up a line, it startled me. took a few more times before I realized what it was that I was typing. |
| 14:24:55 | TomW | I like vi |
| 14:24:58 | RedBeard | I bet it was a supprise |
| 14:25:01 | RedBeard | I do to |
| 14:25:11 | RedBeard | I've not learned emacs |
| 14:25:43 | TomW | I like the fact that I can edit a file without having to move my fingers off the keyboard (as in not reaching for the cursor pad). |
| 14:26:13 | RedBeard | yeah, or the mouse |
| 14:26:38 | TomW | I haven't had a need for emacs. It took me a while to learn vi and I wasn't about to take more time to learn emacs when I got what I needed done in vi. |
| 14:26:57 | RedBeard | that's where I am too |
| 14:27:29 | TomW | Makefiles. learn to write makefiles. |
| 14:27:49 | RedBeard | been learning, I've still got a long way to go though |
| 14:28:03 | TomW | prpplague: finally get awake? |
| 14:28:42 | TomW | prpplague: that is the problem with you westerners, we in the east have been at work for hours already. ;-) |
| 14:29:56 | RedBeard | is in Utah, but has been up for nearly three hours |
| 14:30:15 | TomW | RedBeard: Well, I have to start tracing through that ioremap stuff. later. |
| 14:32:43 | RedBeard | coap |
| 14:32:46 | RedBeard | ciao even |
| 14:33:26 | prpplague | TomW: lol, ya well, been awake for awhile, been occupied with getting my son off to school |
| 14:33:39 | TomW | heh, how old? |
| 14:33:53 | prpplague | 4.5 |
| 14:34:22 | TomW | oh boy, yeah, he doesn't want to be trapped inside of a building all day long. |
| 14:45:40 | prpplague | TomW: its not too bad, he gets to play and learn alot |
| 14:47:01 | TomW | ... and tease the little girls. |
| 14:47:42 | TomW | .. only he hasn't discovered that benefit of going to school yet. :-) |
| 14:48:52 | prpplague | TomW: ya well, he's already interested in "kiss'n the pretty girls" |
| 14:49:12 | prpplague | TomW: i'm afraid we are going to have real problems when he hits puberty |
| 14:49:59 | TomW | heh, so, when is your daughter going to school? |
| 14:51:26 | prpplague | TomW: just a son |
| 14:51:54 | TomW | prpplague: so far.. |
| 14:52:24 | prpplague | TomW: "." , my wife an i can't have kids, we adopted |
| 14:53:45 | TomW | prpplague: ah, ok. maybe you'll have to do the same with a little girl, then you can fret over her meeting some boy like your son. |
| 14:54:18 | prpplague | TomW: lol, ya thats what i'm afraid of already |
| 14:56:40 | TomW | wow! CSCO just popped over $2.50 a share, wonder what their earnings report was last night? |
| 14:57:21 | CosmicPenguin | TomW: The rumor was that their earnings were going to be huge |
| 14:58:20 | TomW | glad I held onto mine, I was seriously thinking of pulling a few thousand off the credit card to buy more when it fell below 13. |
| 14:58:34 | CosmicPenguin | TomW: Their earnings trippled last quarter |
| 14:58:56 | CosmicPenguin | The good news is that they are bringing up some of the other tech stocks.... juniper is up 6% |
| 14:59:40 | CosmicPenguin | MSFT up 6%, Riverstone is up 19% |
| 14:59:48 | prpplague | CosmicPenguin: well i'm sure that most of that is from sales of their crappy voip phone systems, high dollar low quality |
| 15:00:24 | CosmicPenguin | prpplague: I used to work for a cisco competitor - as far as I am concerned, all their stuff is high dollar low quality |
| 15:00:24 | TomW | heh |
| 15:02:06 | TomW | either of you two ever work with ioremap() ? |
| 15:02:25 | CosmicPenguin | TomW: several times.... ioremap_nocache() too ... :) |
| 15:03:50 | prpplague | CosmicPenguin: which company? |
| 15:03:53 | TomW | CosmicPenguin: I am trying to get CS3 to wiggle on the Tux. I tried "IDE = ioremap (0x60000l, 0x1000);" then "IDE [0] = value;" where IDE is "volatile unsigned char * IDE; " |
| 15:05:05 | TomW | CosmicPenguin: according to the SA1100 mem map, CS3 covers the region of physical memory from 0x60000..0x7FFFF. |
| 15:07:12 | TomW | CosmicPenguin: I set MSC1 high word to my shaping values:" MSC1 &= 0x0000ffffl; MSC1 |= 0x44580000l;", but nothing happens. |
| 15:07:47 | TomW | CosmicPenguin: am I missing something? |
| 15:10:47 | prpplague | TomW: did you kill a chicken and spread the blood around the processor? |
| 15:11:38 | TomW | almost! I am about to dive into the kernel source and start tracking the ioremap call down. use a bunch of printk's to see what is going on. |
| 15:14:39 | CosmicPenguin | prpplague: Cabletron |
| 15:15:27 | prpplague | CosmicPenguin: ahh |
| 15:16:01 | prpplague | CosmicPenguin: we have the cisco voip system, and its a pos(nothing to do with Point of Sale) |
| 15:16:01 | CosmicPenguin | TomW: try ioremap_nocache so it doesn't get mapped in the cache - since the variable is volatile, you have to go right out to the metal for each read/write |
| 15:16:39 | TomW | CosmicPenguin: Overall, the approach I am taking looks correct though? |
| 15:17:30 | CosmicPenguin | TomW: It should be - the general procedure for these sorts of things is to take the memory region and remap it |
| 15:17:33 | CosmicPenguin | pretty standard stuff |
| 15:17:43 | TomW | CosmicPenguin: I had done driver work with uClinux, on that stuff we didn't have to worry about MMU protection of hardware. This is the first "real" driver that I am attempting. |
| 15:17:48 | CosmicPenguin | I don't know if the ARM introduces any particular weirdness though |
| 15:18:17 | TomW | you can bet on it, so far, ARM is pretty weird. |
| 15:19:08 | TomW | That concept of using a shift register bit string to shape the CAS signal definately had me smiling at it's simplicity. |
| 15:48:27 | RedBeard | bbl |
| 16:08:10 | | andersee was last seen on #tuxscreen 1 days, 9 hours, 9 minutes and 38 seconds ago, saying: BZFlag: thx [Tue May 7 07:58:32 2002] |
| 16:08:10 | sjhill | ibot: seen andersee |
| 16:08:24 | sjhill | is andersee on vacation? |
| 16:13:47 | pmisfit | is back (gone 34:30:56) |
| 16:13:55 | pmisfit | reading |
| 16:14:15 | pmisfit | sjhill: quite the contrary, wouldn't be surprised if he's spending a lot of time at lineo these days... ;-) |
| 16:15:37 | pmisfit | TomW: and i think you have to use the writel/writeb/writew if you use ioremap. you can't just write to the location. |
| 16:16:01 | CosmicPenguin | pmisfit: no, if you use ioremap, you should be good to go...... |
| 16:16:17 | CosmicPenguin | because it maps it into your memory space |
| 16:17:24 | TomW | I am going down to a lower level to check if I even have CS3 setup to work. I am writing code inside the blob source to wiggle the pin. That is not working either, so this looks like some basic problem here. |
| 16:22:46 | pmisfit | CosmicPenguin: weird. i am sitting in front of an article in this month's embedded linux journal right now, that says ioremap just returns a token, which must be accessed using the writex/readx functions. |
| 16:23:01 | pmisfit | i was surprised too |
| 16:23:17 | CosmicPenguin | pmisfit: thats not my experience... but then again, ELJ is written for non programmers that want to be programmers |
| 16:23:35 | pmisfit | well... that would be me... ;-) |
| 16:23:48 | CosmicPenguin | I guess, in some ways, you should always use write[bwl] - but I always just got a (void *) that can be used as a memory block |
| 16:24:00 | CosmicPenguin | especially for fun stuff like memcpy(), etc.... |
| 16:24:06 | pmisfit | i think they were being fetishistic. the article was about portability. |
| 16:24:44 | CosmicPenguin | pmisfit: that was Greg KH's article, wasn't it |
| 16:24:49 | pmisfit | correct. |
| 16:25:38 | CosmicPenguin | I saw the article, but then somebody stole it before I could finish reading it |
| 16:25:43 | TomW | pmisfit: what page you on? |
| 16:26:40 | pmisfit | correct. |
| 16:26:44 | pmisfit | 28 |
| 16:28:23 | pmisfit | TomW: pp. 28. i enjoyed it, think you'll find it helpful. |
| 16:29:57 | pmisfit | the ioremap stuff is at the end. |
| 16:30:10 | TomW | yeah, listing 4. Nope, I am doing it correctly, this is some ARM-ism at work here. I cannot get the pin to wiggle in blob (C code), time to code some ARM asm in memsetup-sa1100.S and see if I can wiggle it from there. |
| 16:31:09 | CosmicPenguin | TomW: if it didn't wiggle from blob, you can't hold out much hope for the kernel, though, can you? |
| 16:31:16 | pmisfit | have you tried doing it from userspace using dev2mem or some such thing? |
| 16:31:28 | pmisfit | s/dev2mem/devmem2/ |
| 16:31:31 | TomW | The SA1100 developers manual stinks, all it "infers" is that you have to set the MCS1 to shaping values for the pin access + what type of device is out there, then simply access that memory region. |
| 16:31:44 | pmisfit | "simply". heh. |
| 16:31:47 | TomW | pmisfit: no, nothing with dev2mem. |
| 16:43:06 | fontenot | hi |
| 16:43:22 | pmisfit | fontenot: yo |
| 16:43:28 | fontenot | :-) |
| 16:43:37 | sjhill | hey fontenot |
| 16:45:45 | fontenot | checks what day it is. |
| 16:46:59 | fontenot | pmisfit: i saw what you said about the oom killer but it looks like that would only protect the x server - I am having problems with clients and other processes dying too |
| 16:47:28 | fontenot | it definitely looks like you found the right place in the kernel though |
| 16:56:05 | pmisfit | pmisfit: care to try just commenting out the call to oom_kill entirely ;-) ? |
| 16:57:30 | pmisfit | has a talent for unbelievably ham-handed hacks |
| 16:57:48 | pmisfit | fontenot: care to try just commenting out the call to oom_kill entirely ;-) ? |
| 17:10:12 | pattieja | is away: lunch |
| 17:10:40 | fontenot | pmisfit: heh, my ideal solution would have the OOM killer strike at 500kB-1MB free ram rather than 4.5MB |
| 17:10:57 | fontenot | actually my ideal solution would have it be configurable |
| 17:11:01 | fontenot | :-\ |
| 17:30:14 | pmisfit | fontenot: IIRC, the oomkiller goes off when some number of page allocations fail. |
| 17:30:26 | pmisfit | i don't remember it being tied to how much free ram is/isn't present. |
| 17:35:37 | fontenot | pmisfit: then something is really wrong, because I have been unable to get the low water mark below 4MB |
| 17:35:47 | fontenot | even starting small apps |
| 17:38:30 | pmisfit | weird. |
| 17:40:37 | pmisfit | fires up cscope on his kernel sources and has a look |
| 17:41:10 | pmisfit | fontenot: ok, here it is, the function out_of_memory in mm/oom_kill.c |
| 17:41:49 | pmisfit | this is precisely how oomkiller "decides" when to go on a killing spree |
| 17:42:02 | fontenot | cool |
| 17:43:02 | pmisfit | oomkillers don't kill processes, programmers kill processes ;-) |
| 17:44:54 | fontenot | true |
| 17:45:06 | pmisfit | might want to have a look and try changing/commenting-out there. it's pretty clear, well-commented code |
| 17:45:25 | fontenot | ok thanks, but now i have to get ready for work :-\ |
| 17:45:54 | pmisfit | fair enough. i have to get back to work. |
| 17:48:05 | pmisfit | is away: enough slacking already... |
| 17:50:38 | pattieja | is back (gone 00:40:27) |
| 18:27:10 | andersee | CosmicPenguin: you lurking about? |
| 18:36:21 | sjhill | andersee! |
| 18:41:15 | andersee | sjhill: Howdy |
| 18:41:19 | andersee | sjhill: I'm gonna get your patch in today |
| 18:41:56 | jacques | hi andersee |
| 18:42:14 | jacques | any word on locale or wchar? :-) |
| 18:42:38 | andersee | jacques: looked in CVS recently? |
| 18:43:08 | andersee | jacques: went in Monday morning. |
| 18:43:12 | jacques | andersee: nope. why go to that trouble when I have you? :-D |
| 18:43:21 | jacques | andersee: whooo!!!! :-) |
| 18:43:49 | andersee | jacques: there are still a few missing little bits, but most stuff is there now. |
| 18:44:04 | CosmicPenguin | andersee: welcome back in... :) |
| 18:44:05 | sjhill | andersee: excellent...yup, busybox is working great and most importantly the insmod functionality works great on MIPS |
| 18:44:23 | andersee | sjhill: excellent! |
| 18:44:40 | andersee | sjhill: That insmod code was largely untested, so good to know. |
| 18:44:52 | andersee | untested on mips that is. |
| 18:45:27 | CosmicPenguin | andersee: I got a klez e-mail from "you" today.... somebody must have both of us in their outlook directory... :) |
| 18:45:34 | jacques | speaking of busybox... |
| 18:45:44 | jacques | how's modprobe in 0.60.3 ? |
| 18:45:58 | CosmicPenguin | jacques: it needs to support -M for debug... :) |
| 18:46:15 | CosmicPenguin | sorry, insmod does.... |
| 18:46:25 | CosmicPenguin | needs to get wacked over the head with a large object |
| 18:46:40 | sjhill | hands crowbar to CosmicPenguin |
| 18:46:47 | CosmicPenguin | thanks! |
| 18:46:53 | CosmicPenguin | goes into the other room with the crowbar |
| 18:47:01 | andersee | jacques: yeah, the odd thing is that it works perfectly elsewhere. But on the tuxscreen busybox modprobe seems broken. |
| 18:47:01 | sjhill | hehe |
| 18:47:41 | andersee | jacques: I finished writing a paper last night (finally) so I'm hoping to have time to dig into the modprobe problem today. |
| 18:47:57 | andersee | jacques: Of course, if my wife has our baby today all bets are off... |
| 18:48:17 | jacques | andersee: didnt your wife just have a baby like 3 months ago? |
| 18:48:22 | jacques | :-) |
| 18:48:35 | andersee | jacques: I think that was BZFlag |
| 18:48:41 | jacques | heh lol |
| 18:48:55 | jacques | sorry for confusing ppl |
| 18:49:24 | andersee | jacques: I'm working on the 9 women == a baby in one month plan... |
| 18:49:28 | andersee | ;-) |
| 18:50:17 | CosmicPenguin | andersee: yeah, but then you have to pay all 9... it gets expensive |
| 18:50:36 | CosmicPenguin | just think how much 1 costs... :) |
| 18:50:40 | andersee | hehe |
| 18:54:15 | jacques | andersee: if you want nine pregnant women around at the same time I think you are totally nuts. |
| 18:55:56 | andersee | yeah, that would be a bit much |
| 18:58:39 | prpplague | andersee: as i was telling TomW, adoption is the way to go, costs a little more, but then you don't have to deal with your wife 9 months! |
| 19:00:16 | andersee | prpplague: I think the whole human race should switch to a policy of universal adoption. :) |
| 19:46:22 | jacques | does anybody know why sourceforge has started directing downloads to mirrors? |
| 19:46:36 | jacques | oooh 2.1kB/s |
| 19:47:27 | jacques | OH CRAP |
| 19:47:39 | jacques | latest uclibc snap wont compile with my toolchain |
| 19:47:46 | jacques | this is the first time |
| 19:49:26 | andersee | Did you enable locale support? |
| 19:51:00 | jacques | andersee: not that I'm aware of, but this is an internal compiler error |
| 19:51:11 | andersee | oh |
| 19:51:12 | andersee | :( |
| 19:51:29 | jacques | Ive had this problem in the past with arm xcompiler but never on uclibc |
| 19:52:39 | andersee | details? |
| 20:19:24 | | bonjour, mallum_food |
| 20:19:24 | mallum_food | jacques: hi |
| 20:58:49 | jacques | sorry - meeting |
| 20:58:54 | jacques | hi mallum |
| 20:59:04 | jacques | andersee: here are the details |
| 21:03:48 | jacques | In file included from ldso.c:1354: |
| 21:03:49 | jacques | readelflib1.c: In function `_dl_load_elf_shared_library': |
| 21:03:49 | jacques | readelflib1.c:612: internal error--unrecognizable insn: |
| 21:03:49 | jacques | (insn 2270 2267 67 (set (reg:QI 8 r8) |
| 21:03:49 | jacques | (mem:QI (plus:SI (reg:SI 8 r8) |
| 21:03:50 | jacques | (const_int -308 [0xfffffecc])) 0)) -1 (nil) |
| 21:10:04 | andersee | jacques: gcc --version |
| 21:10:21 | andersee | jacques: for your arm compiler... |
| 21:11:49 | jacques | andersee: it was the head of the 2.95 tree a couple months ago + the phil blundell patch |
| 21:11:54 | jacques | I made te toolchain myselg |
| 21:11:58 | jacques | myself |
| 21:13:04 | andersee | hmmm |
| 21:15:15 | jacques | andersee: which toolchain do you use? |
| 21:15:53 | andersee | The debian gcc-2.95 one |
| 21:15:58 | jacques | arg |
| 21:16:01 | jacques | I am not on debian |
| 21:16:10 | jacques | and am not going to convert for this |
| 21:16:25 | andersee | Here are the patches I use |
| 21:16:27 | andersee | http://codepoet.org/toolchain/gcc-2.95/sources/ |
| 21:16:37 | jacques | i guess i have to figure out how to use their bizarre tarballs |
| 21:16:59 | jacques | that's a lot |
| 21:17:00 | andersee | Several arm fixes in there... |
| 21:17:11 | jacques | andersee: how do you apply all those? surely not by hand |
| 21:17:31 | andersee | http://codepoet.org/toolchain/gcc-2.95/Makefile |
| 21:17:41 | jacques | anok thanks |
| 21:17:51 | jacques | looks like I will be making a new toolchain :=/ |
| 21:18:02 | andersee | That is for building a uClibc toolchain, but it also works for glibc toolchains as well. |
| 21:18:19 | andersee | Though the makefile would need adjusting for that |
| 21:18:49 | andersee | That uclibc toolchain just build C, since 2.95.x libstdc++ has issues when not using glibc |
| 21:19:49 | jacques | andersee: ok cool. I only need C |
| 21:21:03 | andersee | Tarballs are at: http://codepoet.org/toolchain/ or uClibc mirrors |
| 21:22:57 | jacques | andersee: where can I find a list of the mirrors? |
| 21:23:12 | andersee | de.busybox.net is fastest. |
| 21:23:49 | andersee | http://de.busybox.net/downloads/uClibc/toolchain/ |
| 21:25:25 | jacques | thanks, downloading |
| 21:27:07 | andersee | jacques: I would guess that gcc-006-arm-pic.patch is the missing bit for you... |
| 21:33:25 | jacques | andersee: hmmm :-) |
| 21:33:29 | jacques | maybe i try just that one |
| 21:34:50 | andersee | dunno. Worth trying to apply all the arm patches I suppose. |
| 21:35:32 | jacques | I'll build toolchains for all permutations of patches and try them all |
| 21:35:37 | jacques | or not |
| 21:36:09 | andersee | heh |
| 21:38:29 | jacques | maybe if I had that dual athlon |
| 22:20:46 | Dark-Work | Yo |
| 22:20:47 | Dark-Work | :) |
| 22:22:15 | Dark-Work | Bonjour ibot ;) |
| 22:35:48 | kergoth | how large is uclibc? |
| 22:39:10 | jacques | built with what options on what platform with what compiler? |
| 22:42:12 | kergoth | heheh. figured |
| 22:42:20 | kergoth | gcc 2.95.3, strongarm |
| 22:42:29 | kergoth | how large is it on the tuxscreen specifically? |
| 22:42:37 | jacques | checking |
| 22:42:41 | jacques | arg |
| 22:42:45 | jacques | cant check right now |
| 22:42:50 | jacques | for various reasons |
| 22:42:54 | jacques | well maybe... |
| 22:42:56 | jacques | just a sec |
| 22:43:28 | kergoth | hehe |
| 22:43:29 | kergoth | no prob |
| 22:44:54 | jacques | -rwxr-xr-x 1 root root 21904 Apr 30 12:50 ld-uClibc-0.9.11.so |
| 22:44:54 | jacques | -rw-r--r-- 1 root root 11528 Apr 30 12:29 libcrypt-0.9.11.so |
| 22:44:54 | jacques | -rw-r--r-- 1 root root 6576 Apr 30 12:29 libdl-0.9.11.so |
| 22:44:54 | jacques | -rw-r--r-- 1 root root 55356 Apr 30 12:29 libm-0.9.11.so |
| 22:44:54 | jacques | -rw-r--r-- 1 root root 74844 Apr 30 12:29 libpthread-0.9.11.so |
| 22:44:55 | jacques | -rw-r--r-- 1 root root 1584 Apr 30 12:29 libresolv-0.9.11.so |
| 22:44:56 | jacques | -rw-r--r-- 1 root root 248852 Apr 30 12:29 libuClibc-0.9.11.so |
| 22:44:58 | jacques | -rw-r--r-- 1 root root 4656 Apr 30 12:29 libutil-0.9.11.so |
| 22:45:14 | kergoth | damn |
| 22:45:21 | kergoth | glibc is massive in comparison |
| 22:45:24 | kergoth | nifty. |
| 22:45:43 | jacques | yes it is |
| 22:45:44 | kergoth | just noticed libm alone in my glibc crosscompile was 500k.. |
| 22:45:45 | kergoth | hehe |
| 22:46:00 | jacques | factor of 10 |
| 22:46:00 | kergoth | thanks for the info |
| 22:46:04 | jacques | for libm |
| 22:46:07 | kergoth | yep |
| 22:46:17 | jacques | you;re welcome |